Auxot exposes two families of API:
- Inference APIs — drop-in compatible with OpenAI and Anthropic SDKs. These are what your applications and tools use to send messages.
- Management API — create and configure agents, providers, skills, and other resources. The web UI and admin agent both use this API.
Use the sidebar to navigate to the specific surface you need.
Base URL
All API endpoints are served from your Auxot instance:
http://localhost:8080
The default bind port is 8080. Set AUXOT_BASE_URL in your environment to your public-facing URL for production.
| Path prefix | Purpose |
|---|---|
/api/openai/ | OpenAI-compatible inference |
/api/anthropic/ | Anthropic-compatible inference |
/api/ | Management API (agents, providers, skills, etc.) |
/v1/intake/ | Webhook intake endpoints for workflow triggers |
Authentication
All endpoints except /api/health and /api/setup/admin require authentication.
Pass your API key in the Authorization header:
Authorization: Bearer user.abc123...
The x-api-key header is also accepted (useful for SDKs that follow the Anthropic convention):
x-api-key: user.abc123...
Both headers work on all routes.
See Authentication & API Keys → for how to create and manage keys.
Error Responses
All errors return a JSON body with an error field and a standard HTTP status code:
{"error": "agent not found"}
Common status codes:
| Status | Meaning |
|---|---|
| 400 | Malformed request or missing required fields |
| 401 | Missing or invalid credentials |
| 403 | Valid credentials but insufficient permissions |
| 404 | Resource not found |
| 409 | Conflict (for example, a resource already exists) |
| 422 | Valid JSON but logically invalid input |
| 500 | Unexpected server error |
Inference APIs
For sending messages to models, see the provider-compatible pages:
These endpoints accept any SDK or HTTP client built for OpenAI or Anthropic. Change the base URL to point at your Auxot instance and use an Auxot API key.